Collaboration framework for modeling

Abstract

A modeling solution, which enables various users to create, use, and evaluate models, such as scientific models, in various scenarios is provided. To this extent, the modeling solution can define a model using one or more model components, each of which is configured to implement a modeling function using a set of standard execution rules. Additionally, the modeling solution can manage attribute data for each model / model component, which can include evaluation data for the corresponding model / model component. A user can review the attribute data and decide on a model / model component to use in a scenario. Subsequently, the user can provide initialization data for the scenario. The scenario can be executed, and result data for the scenario can be stored together with the initialization data. Users also can evaluate the scenario. In an embodiment, multiple users are classified into multiple tiers, and evaluation data is managed for each tier, thereby allowing users to effectively evaluate the model / model component among groups of users with similar concerns.

Description

TECHNICAL FIELD[0001]The disclosure relates generally to modeling, and more particularly, to a collaboration framework for managing models.BACKGROUND ART[0002]Increasingly, modeling, such as scientific modeling, is an important tool for evaluating various systems and assumptions with respect to how these systems behave. Models are commonly used to simulate a system in order to explain observations of the behavior of a system, predict how a system will behave when one or more aspects of the system are changed, etc. Models can be developed and used to explain numerous types of scientific and / or business systems, such as climate, hydrology, geology, epidemiology, economic processes, business processes, etc.[0003]Many models are complex, expensive to develop and / or run, and are made available only to the technical / scientific community that develops each model. Further, a model may only focus on a specific (however complex) set of input data, algorithms, and / or results for a system, whic...

AI Technical Summary

Benefits of technology

[0005]Additionally, the inventors recognize that current implementations of models are inefficient. For example, multiple groups of researchers may be developing models that seek to explain the same system. Each model may include components (e.g., core algorithms) that closely match each other in function (e.g., same or similar input and output data fields), but differ in their implementations and corresponding effectiveness in generating accurate data. Currently, a user must select one model or the other, accepting the limitations of the model selected, or perform multiple simulation runs, each with a different model and attempt to merge the results. A more desirable implementation is to enable the user to utilize the strengths of both models. For example, the user could select the more effective components from two or more models and utilize them together to perform a single simulation run. The resulting simulation run can be both more effective as well as more efficient.

[0007]Aspects of the invention provide a modeling solution, which enables various users to create, use, and evaluate models, such as scientific models, in various scenarios. To this extent, the modeling solution can define a model using one or more model components, each of which is configured to implement a modeling function using a set of standard execution rules. Additionally, the modeling solution can manage attribute data for each model / model component, which can include evaluation data for the corresponding model / model component. A user can review the attribute data and decide on a model / model component to use in a scenario. Subsequently, the user can provide initialization data for the scenario. The scenario can be executed, and result data for the scenario can be stored together with the initialization data. Users also can evaluate the scenario. In an embodiment, multiple users are classified into multiple tiers, and evaluation data is managed for each tier, thereby allowing users to effectively evaluate the model / model component among groups of users with similar concerns.

Problems solved by technology

However, the current implementation framework for scientific models (e.g., limited accessibility, expense, platform-dependency, etc.) greatly restricts access to this information to a small group of “informational elite”.

Since the ability of this informational elite to disseminate the information is limited, uninformed actions may be taken having adverse consequences to the environment.

Additionally, the inventors recognize that current implementations of models are inefficient.

Images